In a dangerous world, the ability to predict safety is paramount. Impairments in predicting safety would lead one to behave as if danger was ever present; while hyperfunction of this predictive capacity would lead one to behave is if danger was nonexistent. Given the shared, positive significance of reward and safety it is hypothesized that reward prediction and safety prediction share a common neural encoding. The basolateral amygdala (BLA) and VS are two brain areas implicated in signaling reward prediction. The objective of the ROO phase of the proposal is to uncover roles for the BLA and VS in safety; learning driven by the omission of aversive events. To this end neurotoxic lesions will be used to describe contributions of the BLA and VS to safety prediction through conditioned inhibition of Pavlovian fear. Single-unit recording will then be used to (1) examine BLA and VS neural activity supporting safety prediction and (2) compare activity supporting safety prediction to activity supporting reward prediction.